DESCRIPTION: (Applicant's Abstract) Natural history of prostate cancer is intrigued by its progression pattern; primary site to lymph node then to bone. Clinical observations indicated that bone may provide a unique microenvironment for the growth of recurrent androgen independent (AI) prostate cancer. Mortality of prostate cancer patients is due to the emergence of AI prostate cancer cells from bony metastatic site. No effective treatment has been applied because not many AI models are available for studying the underlying mechanisms. Recently, the applicant established a chimeric tumor model by coinoculating human prostate cancer cell line (LNCaP) and bone stromal cells (MS) into athymic nude mouse where majority of tumors appeared to be carcinoma. And, prostate-specific antigen (PSA) levels, a tissue-specific protein secreted by a prostatic epithelium exclusively, detected from tumor bearing animals, correlated with the volume of prostate tumors and further upregulated during AI progression. Also, the applicant demonstrated that bone stromal cell is able to facilitate the emergence of AI prostate cancer. Taken together, data indicated that the LNCaP tumor model mimics clinical observations from prostate cancer progression. Cell adhesion molecules (CAMs) are known to play a critical role in cell differentiation and development. Moreover, data from molecular genetic studies suggest that CAMs may play a suppressive role in tumorigenesis. Recently the applicant demonstrated that C-CAM, an epithelial-specific CAM regulated by androgen acts as a tumor suppressor in prostate cancer progression. Based on these results, he decided to further examine whether C-CAM-expressing recombinant adenovirus could be a potential agent for prostate cancer gene therapy. From in vitro characterization, he observed that majority of human prostate cancer cells are very sensitive to the C-CAM adenoviral infection and that C-CAM mRNA and protein levels can still be detected 20 days after infection, suggesting that the prolonged expression of C- CAM in viral-infected cells may compensate for one of the disadvantages of the recombinant adenovirus with a short half-life. In vivo administration of a single dose of C-CAM adenovirus into the preexisting PC-3 tumors, the AI tumors, was able to suppress the growth of tumors for over 3 weeks. In contrast, the control virus failed to have the same effects. On the other hand, a single dose of C-CAM adenoviral administration could also prevent the recurrent AI tumors in the Shionogi model. Based on these results, the applicant believes that C-CAM adenovirus can be a potential agent for prostate cancer therapy. In this study, the applicant proposes to determine the optimal dose and treatment schedule of C- CAM adenovirus for prostate cancer using both the PC-3 and Shionogi models. He also would like to examine the effect of C-CAM on the intervention of epithelial-stromal interaction using the LNCaP chimeric model for prostate cancer progression. Finally, he will take advantage of the tissue-specific expression of PSA promoter to design a novel therapeutic strategy in prostate cancer therapy.